A Season of Simulations - NFL Python Sim

Danchat
Content Creator

Posts: 19,021

Likes: 16,601

A Season of Simulations - NFL Python Sim Feb 22, 2022 22:56:05 GMT -6 Reignman likes this

Quote

Post by Danchat on Feb 22, 2022 22:56:05 GMT -6

Reignman said:

What kind of variables are involved in this simulation?

A lot! I'll go more into depth later, as this thread just covers the basics.

Feb 22, 2022 22:15:10 GMT -6 Reignman said:

How does it know when to go for 2 or call a timeout?

Going for 2 is based on game theory, so when leading by 1, 5, 12, or 19 points, and so on. Every team will always attempt the extra 2 in these situations, which isn't 100% realistic.

Taking timeouts is tricky, at least when the offense is taking them. Basically, if the clock has to run and they've got the ball, they'll take a timeout. This can lead to teams blowing all 3 of their timeouts in the first 3 plays within 2:00, but the alternative is teams screwing up the final seconds of the game and not being able to precisely call them at the right time. Timeouts will also be taken if the offense is trying to run out the clock at the end of the 4th quarter, which was pretty simple to program.

If you have suggestions on how to improve the code, I'm all ears!

Feb 22, 2022 22:15:10 GMT -6 Reignman said:

How does it know how much time to take off the clock for each play?

There is a base 'play length' value for every type of play, takes more with plays that go for more yards, and then a random number of seconds is added for the sake of randomness.

Feb 22, 2022 22:15:10 GMT -6 Reignman said:

I didn't go through and tally up his yards, but did Joe get the extra 9 yards here when he scored from the 32 lol? And did the defense get a late hit penalty for tackling him 9 yards deep xD?

It was a 32 yard TD and it was scored as that, the way I report it tells you that it was going to be a 41 yard run if he hadn't reached the endzone. That is something I will patch up later, but I like being able to see how many yards the play was going to go if it hadn't ended in a TD.

Feb 22, 2022 22:15:10 GMT -6 Reignman said:

Oooo time to bust the simulator. How good would a 200 rated player be? Oh wait, you'd have to disable that pesky min/max check thing *whistle*

. In Madden I can get the PLAY file to accept ratings above 99, but they still simulate like 99, which is unfortunate when you just want to mess around xD.

I've coded the player rating conversions to max out at a certain rate, so you're not going to be able to break the app that way!

However, you can test the "running leads to winning" philosophy all you like by setting the RunPassRatio to 100. This won't result in 100% of run plays being called, but it'll be crazy. Or you could try having 100% of the targets go to Justin Jefferson... the possibilities are endless!

Last Edit: Feb 22, 2022 22:57:16 GMT -6 by Danchat

Danchat
Content Creator

Posts: 19,021

Likes: 16,601

A Season of Simulations - NFL Python Sim Feb 23, 2022 23:40:02 GMT -6

Quote

Post by Danchat on Feb 23, 2022 23:40:02 GMT -6

Feb 22, 2022 22:15:10 GMT -6 Reignman said:

QB accuracy seems to be a bit off too.

Missed this one - over a few thousand simulations, the average QB accuracy is around 60%. I want to get that up to about 64%, but I have to do it without breaking the other stats like YPA and passing yards per game.

Reignman
Purple Pain Lap Dog

Posts: 44,727

Likes: 30,666
Fan Since: 1984

A Season of Simulations - NFL Python Sim Mar 4, 2022 22:52:24 GMT -6

Quote

Post by Reignman on Mar 4, 2022 22:52:24 GMT -6

Feb 22, 2022 22:56:05 GMT -6 Danchat said:

If you have suggestions on how to improve the code, I'm all ears!

I might come up with some ideas. You say you use the PLAY database from Madden, what about coaching? I don't necessarily like the coaching attributes in Madden because most of those settings have no bearing on simulated games, so what about your own coaching profiles? You probably wouldn't need one for every single coach, just some basic generic ones, but with the ability to add/use more if you want to try something specific, like Brandon Staley going for it on every 4th down xD. Or what would a game look like with no punting. Yeah you could really some analytics to the test here. You can't really do stuff like that in Madden. Simulate a season with no punting, only 2 point conversions, or no passing xD.

Feb 22, 2022 22:56:05 GMT -6 Danchat said:

That is something I will patch up later, but I like being able to see how many yards the play was going to go if it hadn't ended in a TD.

In that case I bet Bo Jackson is still running down Hwy 5 when he ran out of the Kingdome tunnel back in 1987 *lol*

Feb 22, 2022 22:56:05 GMT -6 Danchat said:

I've coded the player rating conversions to max out at a certain rate, so you're not going to be able to break the app that way!

Well that's no fun. We need a "nothing's safe" Reignman version. *whistle*

Feb 22, 2022 22:56:05 GMT -6 Danchat said:

However, you can test the "running leads to winning" philosophy all you like by setting the RunPassRatio to 100. This won't result in 100% of run plays being called, but it'll be crazy. Or you could try having 100% of the targets go to Justin Jefferson... the possibilities are endless!

Now that's more like it.

Aug 8, 2024 18:42:12 GMT -6 whoskmoon said:

if sitting and watching something versus actually doing it made a person great at it, Reignman would be the next Ron Jeremy.

Raptorman
Hall of Fame Inductee

Posts: 18,510

Likes: 8,265
Fan Since: November 6, 1966.
Member is Online

A Season of Simulations - NFL Python Sim Mar 5, 2022 11:24:11 GMT -6

Quote

Post by Raptorman on Mar 5, 2022 11:24:11 GMT -6

So, are you going to update the roster and then run simulations of games during the year and post them before the games? I would love to see how close it gets. Of course programming intangibles into it would be tough.

Also, QB accuracy is around 66% in the NFL if I am correct, why 64%?

You can't truly call yourself "peaceful" unless you are capable of great violence. If you are not capable of great violence, you are not peaceful, you are harmless.

Danchat
Content Creator

Posts: 19,021

Likes: 16,601

A Season of Simulations - NFL Python Sim Mar 5, 2022 12:08:09 GMT -6

Quote

Post by Danchat on Mar 5, 2022 12:08:09 GMT -6

I'm currently working on a web app for the simulation, and I am also revamping the code and working on getting penalties to work properly. I'll continue to post updates for how the 2021 season re-simmed is going for the Vikings.

Can't remember if I mentioned this before, but I simmed the Super Bowl matchup 100 times, and the Rams won 53 times.

Mar 4, 2022 22:52:24 GMT -6 Reignman said:

what about coaching? I don't necessarily like the coaching attributes in Madden because most of those settings have no bearing on simulated games, so what about your own coaching profiles? You probably wouldn't need one for every single coach, just some basic generic ones, but with the ability to add/use more if you want to try something specific, like Brandon Staley going for it on every 4th down xD. Or what would a game look like with no punting. Yeah you could really some analytics to the test here. You can't really do stuff like that in Madden. Simulate a season with no punting, only 2 point conversions, or no passing xD.

I do have a coach table that affects how often teams run the ball, and use their RB/WR/TE/FBs, but I haven't delved into any of the game theory yet. I would like to add '4th down rate' to that table. I don't know how I would do the 2 point conversions, would I make those coaches only go for the obvious 2 point scenarios?

Mar 5, 2022 11:24:11 GMT -6 Raptorman said:

So, are you going to update the roster and then run simulations of games during the year and post them before the games? I would love to see how close it gets. Of course programming intangibles into it would be tough.

Yeah, I'm planning to do something like that for the 2022 season. The intention for the simulation isn't to predict future matchups - there are no factors for home field advantage, weather, intangibles - it's to play a game in a vacuum and mirror how actual NFL games look. But I will definitely be running tests to see how close it can get to mirroring real life.

Mar 5, 2022 11:24:11 GMT -6 Raptorman said:

Also, QB accuracy is around 66% in the NFL if I am correct, why 64%?

I was trying to get the stats to match the 2018-21 seasons on a whole, which I believe came out to around 64%. But it is very dependent on how many accurate QBs are playing - the NFL generally has the most accurate QBs playing at all times, but my script has to be able to account for all sorts of QBs with varying degrees of accuracy. Did that make any sense?

Raptorman
Hall of Fame Inductee

Posts: 18,510

Likes: 8,265
Fan Since: November 6, 1966.
Member is Online

A Season of Simulations - NFL Python Sim Mar 5, 2022 13:03:48 GMT -6

Quote

Post by Raptorman on Mar 5, 2022 13:03:48 GMT -6

Mar 5, 2022 12:08:09 GMT -6 Danchat said:

I'm currently working on a web app for the simulation, and I am also revamping the code and working on getting penalties to work properly. I'll continue to post updates for how the 2021 season re-simmed is going for the Vikings.

Can't remember if I mentioned this before, but I simmed the Super Bowl matchup 100 times, and the Rams won 53 times.

Mar 4, 2022 22:52:24 GMT -6 Reignman said:

what about coaching? I don't necessarily like the coaching attributes in Madden because most of those settings have no bearing on simulated games, so what about your own coaching profiles? You probably wouldn't need one for every single coach, just some basic generic ones, but with the ability to add/use more if you want to try something specific, like Brandon Staley going for it on every 4th down xD. Or what would a game look like with no punting. Yeah you could really some analytics to the test here. You can't really do stuff like that in Madden. Simulate a season with no punting, only 2 point conversions, or no passing xD.

I do have a coach table that affects how often teams run the ball, and use their RB/WR/TE/FBs, but I haven't delved into any of the game theory yet. I would like to add '4th down rate' to that table. I don't know how I would do the 2 point conversions, would I make those coaches only go for the obvious 2 point scenarios?

Mar 5, 2022 11:24:11 GMT -6 Raptorman said:

So, are you going to update the roster and then run simulations of games during the year and post them before the games? I would love to see how close it gets. Of course programming intangibles into it would be tough.

Yeah, I'm planning to do something like that for the 2022 season. The intention for the simulation isn't to predict future matchups - there are no factors for home field advantage, weather, intangibles - it's to play a game in a vacuum and mirror how actual NFL games look. But I will definitely be running tests to see how close it can get to mirroring real life.

Mar 5, 2022 11:24:11 GMT -6 Raptorman said:

Also, QB accuracy is around 66% in the NFL if I am correct, why 64%?

I was trying to get the stats to match the 2018-21 seasons on a whole, which I believe came out to around 64%. But it is very dependent on how many accurate QBs are playing - the NFL generally has the most accurate QBs playing at all times, but my script has to be able to account for all sorts of QBs with varying degrees of accuracy. Did that make any sense?

Clear as mud. No, I get what you are saying.

You can't truly call yourself "peaceful" unless you are capable of great violence. If you are not capable of great violence, you are not peaceful, you are harmless.

Reignman
Purple Pain Lap Dog

Posts: 44,727

Likes: 30,666
Fan Since: 1984

A Season of Simulations - NFL Python Sim Mar 6, 2022 3:35:48 GMT -6

Quote

Post by Reignman on Mar 6, 2022 3:35:48 GMT -6

Mar 5, 2022 12:08:09 GMT -6 Danchat said:

I don't know how I would do the 2 point conversions, would I make those coaches only go for the obvious 2 point scenarios?

What about a sliding rate? It could factor in how much time is remaining in the game. Some coaches start chasing points earlier than others. If you get penalties working, then teams seem to like going for 2 early in the game if they get to attempt them at the 1 yard line instead of the 2.

Once you get everything ironed out with the sim version, you'll have to create a turn by turn version, or a coaches mode where a user can call the plays for one or both teams. I don't know how timeouts would work in that version though. It could get annoying if it had to ask you after every play lol. Timeout? Y/N. Oh wow, what about an actual working clock/play clock? *woot*

And by call plays I mean it doesn't have to be too complicated, just your basic pass, run, punt, FG, 2 point attempt, onside kick, decline penalties. Don't need an actual playbook or anything. The sim would use it's own programming to determine exactly what type of pass to throw on 3rd and long for example. Unless a hail mary is needed maybe.

Another thing to consider is a config file which could act sorta like the global sliders in Madden. A universal modifier to quickly adjust things to the users liking. Maybe they want 90% completion rates and games that end 63-56 lol. Aggressive defenses that produce more sacks and turnovers but give up bigger plays. I better stop now because I got all kinds of ideas flooding in lol. Oh man, now I'm thinking about a graphical version with drive charts and mini helmets xD. Ok, that's too much meth for tonight.

Aug 8, 2024 18:42:12 GMT -6 whoskmoon said:

if sitting and watching something versus actually doing it made a person great at it, Reignman would be the next Ron Jeremy.

Danchat
Content Creator

Posts: 19,021

Likes: 16,601

A Season of Simulations - NFL Python Sim Mar 6, 2022 13:25:23 GMT -6

Quote

Post by Danchat on Mar 6, 2022 13:25:23 GMT -6

Mar 6, 2022 3:35:48 GMT -6 Reignman said:

What about a sliding rate? It could factor in how much time is remaining in the game. Some coaches start chasing points earlier than others. If you get penalties working, then teams seem to like going for 2 early in the game if they get to attempt them at the 1 yard line instead of the 2.

That makes sense, though I don't foresee that happening very often.

Mar 6, 2022 3:35:48 GMT -6 Reignman said:

Once you get everything ironed out with the sim version, you'll have to create a turn by turn version, or a coaches mode where a user can call the plays for one or both teams. I don't know how timeouts would work in that version though. It could get annoying if it had to ask you after every play lol. Timeout? Y/N. Oh wow, what about an actual working clock/play clock?

Eh, I don't know about that. There have been Strat-o-Matic Football games that do this same thing (I bought one a good 12-13 years ago, and even I couldn't handle how dull it was). I have a lot of things on my to-do list - potentially re-creating Madden 08's franchise mode is more the path I'd go down. Boy, imagine that, an open source Madden game? EA would be DMCA'ing that as fast as humanly possible.

Mar 6, 2022 3:35:48 GMT -6 Reignman said:

Another thing to consider is a config file which could act sorta like the global sliders in Madden. A universal modifier to quickly adjust things to the users liking. Maybe they want 90% completion rates and games that end 63-56 lol. Aggressive defenses that produce more sacks and turnovers but give up bigger plays. I better stop now because I got all kinds of ideas flooding in lol. Oh man, now I'm thinking about a graphical version with drive charts and mini helmets xD. Ok, that's too much meth for tonight.

If you want to get on the chart making, sure! There will be plenty of data to pour over, I can assure you of that.

Raptorman Hall of Fame Inductee Posts: 18,510 Likes: 8,265 Fan Since: November 6, 1966. Member is Online	A Season of Simulations - NFL Python Sim Mar 6, 2022 21:28:24 GMT -6 Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by Raptorman on Mar 6, 2022 21:28:24 GMT -6 Reignman and Danchat, maybe you two should get a room?
	You can't truly call yourself "peaceful" unless you are capable of great violence. If you are not capable of great violence, you are not peaceful, you are harmless.

Reignman
Purple Pain Lap Dog

Posts: 44,727

Likes: 30,666
Fan Since: 1984

A Season of Simulations - NFL Python Sim Mar 6, 2022 23:50:13 GMT -6

Quote

Post by Reignman on Mar 6, 2022 23:50:13 GMT -6

Mar 6, 2022 13:25:23 GMT -6 Danchat said:

That makes sense, though I don't foresee that happening very often.

Do you want to be better than EA or not? Their slogan is "if it's in the game, it's in the game", which is a lie.

Mar 6, 2022 13:25:23 GMT -6 Danchat said:

Eh, I don't know about that. There have been Strat-o-Matic Football games that do this same thing

Do you want to be rich or not? You could put strat-o-matic to shame. It could be both a simulator and input based. Maybe you only want to call plays in the 4th quarter or something and simulate up to that point.

Mar 6, 2022 13:25:23 GMT -6 Danchat said:

and even I couldn't handle how dull it was

Impossible!

Mar 6, 2022 13:25:23 GMT -6 Danchat said:

potentially re-creating Madden 08's franchise mode is more the path I'd go down.

I thought that was the idea with this thing? Build a better simulator than Madden. The Madden simulator ignores so many variables.

Mar 6, 2022 13:25:23 GMT -6 Danchat said:

If you want to get on the chart making, sure! There will be plenty of data to pour over, I can assure you of that.

Well if you want a graphical version someday, give me dimensions and an idea of how you'd want it to look.

Mar 6, 2022 21:28:24 GMT -6 Raptorman said:

Reignman and Danchat, maybe you two should get a room?

Pffff maybe we will, where we'll work on our billion dollar football simulator. I'm already thinking about version 4 (season mode), and version 5 where we implement a salary cap/salaries *woot*

. Might have to bring @aussie61 on board for that one.

Aug 8, 2024 18:42:12 GMT -6 whoskmoon said:

if sitting and watching something versus actually doing it made a person great at it, Reignman would be the next Ron Jeremy.

Deleted Ghosted Member Posts: 0 Likes:	A Season of Simulations - NFL Python Sim Mar 7, 2022 2:50:10 GMT -6 comet52 likes this Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by Deleted on Mar 7, 2022 2:50:10 GMT -6 Mar 6, 2022 23:50:13 GMT -6 Reignman said: Might have to bring Aussie on board for that one. I don't understand anything in this thread, so you might want to rethink that

Danchat
Content Creator

Posts: 19,021

Likes: 16,601

A Season of Simulations - NFL Python Sim Mar 7, 2022 14:03:05 GMT -6

Quote

Post by Danchat on Mar 7, 2022 14:03:05 GMT -6

Mar 6, 2022 23:50:13 GMT -6 Reignman said:

Do you want to be rich or not? You could put strat-o-matic to shame. It could be both a simulator and input based. Maybe you only want to call plays in the 4th quarter or something and simulate up to that point.

Where is all this money going to come from? If this is headed down the Madden path, then we aren't getting paid for this for several reasons - I said 'open source' for a reason. The path paved in gold would be using the simulator to predict future games for football betting... but that's not the path I want to take, and the simulator's also too dependent on Madden ratings for that to work well.

Reignman
Purple Pain Lap Dog

Posts: 44,727

Likes: 30,666
Fan Since: 1984

A Season of Simulations - NFL Python Sim Mar 7, 2022 18:59:35 GMT -6

Quote

Post by Reignman on Mar 7, 2022 18:59:35 GMT -6

Mar 7, 2022 2:50:10 GMT -6 @aussie61 said:

Mar 6, 2022 23:50:13 GMT -6 Reignman said:

Might have to bring Aussie on board for that one.

I don't understand anything in this thread, so you might want to rethink that

You don't need to understand any of it, you just need to understand the salary cap. *lol*

What does it take to get a little teamwork around here?

Mar 7, 2022 14:03:05 GMT -6 Danchat said:

Where is all this money going to come from? If this is headed down the Madden path, then we aren't getting paid for this for several reasons - I said 'open source' for a reason. The path paved in gold would be using the simulator to predict future games for football betting... but that's not the path I want to take, and the simulator's also too dependent on Madden ratings for that to work well.

You're not thinking long term enough, or big picture. Yes it'll be the ultimate simulator. I'm already onto version 6, era based settings xD. What would a simulated game look like between the 2007 Patriots and '85 Bears if it were played in 1972 lol? We could even predict where things are headed and simulate it based on what football might be like in 2030. If bettors want to use it to try and predict things, that's up to them. You just want to create the best simulator possible.

Everything will be based on numbers obviously, so we just need data, the more the better, and I have data. With the data you can see tendencies, trends, and patterns, and you can create config files based on all that. Have the sim run all the math, and boom, simulated football games. Throw in some modifiers to allow for more customizing and fine tuning to spice things up. People like having more control so they can try their own thing. I do anyway. I hate when games impose limitations on things because they want you to play/use it the way they think it's intended. Nonsense, you make it, and I'll figure out how I want to use it lol.

I'll show you the 2 point data I'm working on soon. Like hey, here's a fun fact about 2 point conversions. Since 2015 (when PAT's were moved back to the 15), going for 2 has only netted the league a whole 15 points. The NFL has converted 391/810 (48.3%) 2 pointers in that time, and it's converted on 8122/8579 (94.7%) PAT's. So that's 782 points off of 2 pointers alone, but had they kicked the PAT's instead, and converted 94.7% of those 810 attempts, it would have led to 767 points. Obviously you can't just settle for a PAT when you're down by 2 late, but from an analytics perspective, there's really no benefit of going for 2 ... Brandon Staley.

Unless you run on the 2 pointer lol. Last year the NFL only converted 50/109 (45.9%) when passing for 2, but 25/45 (55.6%) when running. The trend has held up since 1998, where it's 44.3% to 54.3% in favor of running.

And with that I just spotted an exploit in the simulator. *skeptic*

We're going to need AI learning in version 7. If you spammed run on every 2 point attempt that would give you an overall advantage, so there should be diminishing returns if you spam the same thing over and over. We're going to need more math.

Aug 8, 2024 18:42:12 GMT -6 whoskmoon said:

if sitting and watching something versus actually doing it made a person great at it, Reignman would be the next Ron Jeremy.

Deleted Ghosted Member Posts: 0 Likes:	A Season of Simulations - NFL Python Sim Mar 7, 2022 19:06:13 GMT -6 Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by Deleted on Mar 7, 2022 19:06:13 GMT -6 Mar 7, 2022 18:59:35 GMT -6 Reignman said: What does it take to get a little teamwork around here? Says the Orange striped defector

Reignman
Purple Pain Lap Dog

Posts: 44,727

Likes: 30,666
Fan Since: 1984

A Season of Simulations - NFL Python Sim Mar 7, 2022 19:39:55 GMT -6

Quote

Post by Reignman on Mar 7, 2022 19:39:55 GMT -6

So here's when teams typically go for 2 throughout a game by the minute, regardless of the math. This is in raw numbers, so those shorter peaks in the first quarter represent about only 1 or 2 attempts, and that taller red spike late in the 4th is 194 attempts. The red line is every attempt since 1998 and the blue line represents the past 5 years. They're basically identical, so the trend hasn't changed, just the frequency. I don't see anything surprising, except maybe that spike just before half lol. The point is, these are the numbers the simulator could take into consideration.

But now lets do it by the math. So here in the left column you see how big the deficit or lead is, then you see how many TD's were scored at that margin, followed by 2 point attempts. Broken down by quarter to give an idea of when teams start chasing points. I'm actually surprised to see that half the teams start chasing points in the 2nd quarter already when they trail by 8, which might explain that end of half spike in the chart above.

          1ST QUARTER            2ND QUARTER            3RD QUARTER            4TH QUARTER
DEFICIT  TD  2PT    PCT         TD  2PT    PCT         TD  2PT    PCT         TD  2PT    PCT
 -1       9    0   0.00        112   13  11.61        112   34  30.36        146  134  91.78
 -2       8    1  12.50         39    5  12.82         57   15  26.32         79   51  64.56
 -3     443    2   0.45        665    3   0.45        297    0   0.00        416    3   0.72
 -4      75    0   0.00        451    7   1.55        278    3   1.08        342    4   1.17
 -5       3    0   0.00         20    7  35.00         52   29  55.77        147  137  93.20
 -6      35    0   0.00        190    0   0.00        143    2   1.40        227    2   0.88
 -7     653    0   0.00        833    8   0.96        319    1   0.31        483   25   5.18
 -8       3    1  33.33         70   32  45.71         92   53  57.61        174  170  97.70
 -9       5    0   0.00         38    0   0.00         42    0   0.00        133    0   0.00
-10      56    2   3.57        347    2   0.58        274    1   0.36        343    0   0.00
-11      10    0   0.00        157   10   6.37        153   49  32.03        201  182  90.55
-12       2    0   0.00         22    2   9.09         31    0   0.00        127    3   2.36
-13       3    0   0.00         94    0   0.00        124    0   0.00        177    1   0.56
-14      52    1   1.92        310    1   0.32        207    2   0.97        363   18   4.96
-15       0    0                15    2  13.33         39    6  15.38        115   18  15.65
-16       1    0   0.00         19    3  15.79         33   20  60.61        113  110  97.35
-17       3    0   0.00        128    0   0.00        122    5   4.10        253    8   3.16
-18       0    0                40    1   2.50         76   11  14.47        132   55  41.67
-19       0    0                 7    2  28.57         24    5  20.83         68   57  83.82
-20       0    0                23    1   4.35         52    1   1.92        106    2   1.89
-21       5    0   0.00         77    0   0.00         96    0   0.00        146   10   6.85
-22       0    0                 1    0   0.00         18    6  33.33         36   15  41.67
-23       0    0                 4    0   0.00         25    5  20.00         57   17  29.82
-24       0    0                25    2   8.00         51   14  27.45        113   78  69.03

          1ST QUARTER            2ND QUARTER            3RD QUARTER            4TH QUARTER
LEAD     TD  2PT    PCT         TD  2PT    PCT         TD  2PT    PCT         TD  2PT    PCT
 1        8    0   0.00        157    3   1.91        121    2   1.65        101    5   4.95
 2       16    0   0.00         27    2   7.41         55    1   1.82         94    1   1.06
 3      284    3   1.06        590    4   0.68        334    0   0.00        302    2   0.66
 4       86    0   0.00        468    2   0.43        237    0   0.00        176    2   1.14
 5        5    0   0.00         29    2   6.90         49    1   2.04         88    4   4.55
 6       22    2   9.09        155   12   7.74        145   25  17.24        112   75  66.96
 7      538    0   0.00        784    4   0.51        329    1   0.30        298    0   0.00
 8        2    0   0.00         47    1   2.13         96    4   4.17        103    3   2.91
 9        4    0   0.00         37    0   0.00         51    2   3.92         53    3   5.66
10       44    1   2.27        329    1   0.30        236    0   0.00        226    1   0.44
11        2    0   0.00        181    2   1.10        163    0   0.00        134    1   0.75
12        0    0                12    0   0.00         29    1   3.45         57    0   0.00
13        1    0   0.00         71    8  11.27         93   10  10.75        109   24  22.02
14       51    0   0.00        288    2   0.69        181    1   0.55        165    3   1.82
15        1    0   0.00         12    0   0.00         38    0   0.00         53    0   0.00
16        0    0                12    2  16.67         30    2   6.67         47    3   6.38
17        2    0   0.00         95    0   0.00        149    0   0.00        111    0   0.00
18        0    0                35    1   2.86         78    0   0.00         67    0   0.00
19        1    0   0.00          3    0   0.00         15    1   6.67         28    2   7.14
20        0    0                18    1   5.56         48    2   4.17         62    1   1.61
21        4    0   0.00         76    0   0.00         95    0   0.00        102    1   0.98
22        0    0                 1    0   0.00          9    0   0.00         32    0   0.00
23        0    0                 3    0   0.00         11    0   0.00         24    0   0.00
24        0    0                16    0   0.00         69    0   0.00         52    0   0.00

tied   3320   20   0.60       1209    8   0.66        407    4   0.98        360    3   0.83